The long-term objective is to develop an immunotherapy for patients with cancer for whom tumor antigens remain unidentified. The systemic administration of dendritic cells (DC) that have been pulsed/fused ex vivo with tumor antigens has demonstrated considerable promise as a strategy for enhancing antitumor immune responses. An alternative strategy is to inject DC directly into tumors, where they can process the tumor antigens present. A pilot study of the intratumoral (IT) administration of DC has been completed in patients with metastatic cancer. Therapy was technically expedient, well tolerated, and resulted in regression of the injected tumor associated with an infiltration of T-cells. Although the DC were active in the tumors injected, there was little evidence of systemic antitumor immunity. The induction of systemic immunity with IT DC may be regulated at the level of the responding T cell and the provision of adequate T-helper (Th) 1 activity. How to most effectively provide Thl activity is not known. The dose and schedule by which Th1 cytokines, such as interleukin - (IL-) 2, are administered may influence efficacy. The T cells infiltrating DC-injected tumors in the pilot study reacted with heat shock proteins (HSP) derived from the tumor. HSP function in antigen transport. The role of HSP in antitumor immune responses in humans is also not known. Thus, a pilot clinical trial to examine how systemic IL-2 modulates the induction of antitumor immunity with IT DC and the role of HSP will be performed in patients with melanoma. The specific aims are: 1) to examine the ability of IT DC plus low-dose IL-2 to generate systemic immune responses to tumor-derived HSP and 2) to evaluate sequential versus concurrent IL-2 in order to select a regimen for further study. Consenting, eligible patients with metastatic melanoma will be enrolled either into a sequential-IL-2 or into a concurrent-IL-2 treatment cohort. Patients will receive low-dose IL-2 by subcutaneous injection either two weeks after completing three IT injections of their DC or with each IT injection. The injected tumor and an uninjected control tumor will be excised to examine immune responses. Systemic cellular and humoral immunity to HSP derived from the patients own tumor will be assessed. IT immunologic activity and migration of the DC will also be assessed to examine the mechanisms involved and also to serve as controls. The results of this study may lead to a practical and non-toxic cancer immunotherapy, improve the application of IL-2 as an immunologic adjuvant, help establish the clinical utility of assessing immune responses to HSP, and lead to the discovery of unique tumor antigens.